In the field of a smoothing process for smoothing image data to obtain easily viewable image data, an LPF (Low Pass Filter) that is a filter for passing only a low-pass frequency has been conventionally utilized. In general, when an LPF process is performed, it is necessary to hold pixels in a smoothing range and thus to prepare a memory that has a capacity for the number of lines required to hold the pixels as illustrated in FIG. 12A. For example, in FIG. 12A, because pixels in the range of a gray part should be held with respect to target pixels, the number of lines depicted by thick arrows is stored in a memory. Therefore, a memory having a large capacity is required.
In recent years, a technology for reducing a memory capacity necessary for holding pixels within a smoothing process range has been known as disclosed in, for example, Japanese Laid-open Patent Publication No. 2004-221645. As illustrated in FIG. 12B, according to this technology, a reduced image obtained by reducing an image before one frame is held and a smoothing process is performed by using the reduced image. When the reduction rate of the reduced images is adequately small, the smoothing process can be performed by a memory that has a capacity smaller than a memory that has a capacity for the number of lines as described above.
However, there is a problem in that edges cannot be accurately reproduced by the LPF process in the conventional art. Specifically, because the fine information of an image is lost by an image reduction process, edges cannot be accurately reproduced.